This work provides an explanatory analysis of the influence of input parameters on the performance of subspace-based Direction of Arrival (DoA) estimation algorithms. The objective of this work is twofold. First, to drive a Steering Vector (SV) that works for arbitrary array configuration rather than just Uniform Linear Array (ULA) geometry. Second, to identify how the performance of the subspace-based algorithms is affected by tuning the input parameters. The later objective is crucial as it allows optimizing the algorithm through selecting optimum parameters to set an appropriate tradeoff between complexity and performance based on the intended applications. Toward that end, we firstly drive an SV for arbitrary array configuration followed by revealing the working principle of subspace based DoA techniques. Secondly, we evaluate the impact of several parameters namely Signal to Noise Ratio (SNR), number of snapshots, number of array elements, separation between array elements, number of available sources, and dependency between sources to conduct our analysis. Numerical simulations over a wide range of scenarios along with intensive Monte Carlo simulations are conducted to show the influence of these parameters on the resolution, accuracy, and complexity of the subspace based DoA estimation algorithm. As demonstrated by the obtained results, the performance of this class of DoA estimation method is determined mostly by the values of the input parameters. Furthermore, the simulation results show that tradeoff between performance and computational complexity needs to be considered when the system parameters are chosen for DoA estimation algorithms.